Fixes #18781: Replace stats API by a prometheus endpoint

relay/sources/api-doc/code_samples/curl/metrics.sh

@@ -0,0 +1 @@
+curl http://localhost:3030/metrics

relay/sources/api-doc/introduction.md

@@ -4,9 +4,13 @@ Download OpenAPI specification: [openapi.yml](openapi.yml)
 
 Rudder relay exposes an internal API, enabling the agents and other services to interact with the relay.
 
+A part of the API (everything but the `System` category) is exposed publicly on the relay
+at `https://relay.example.com/rudder/relay-api/` (which is a reverse proxy of
+`http://localhost:3030/rudder/relay-api/1/`).
+
 ## Authentication
 
-The relay API is authentified on a case-by-case basis:
+The relay API is authenticated on a case-by-case basis:
 
 * *Remote run* is only accessible to requests from the IP of the relay's own policy server
 

relay/sources/api-doc/openapi.src.yml

@@ -15,10 +15,8 @@ info:
     name: "CC-BY-SA 2.0"
     url: "https://spdx.org/licenses/CC-BY-SA-2.0.html"
 servers:
-  - url: "http://localhost:3030/rudder/relay-api/1/"
+  - url: "http://localhost:3030"
     description: "Local API"
-  - url: "https://rudder.example.local/rudder/relay-api/"
-    description: "Public API (does not include the `System` API)"
 externalDocs:
   description: Learn more about Rudder.
   url: "https://docs.rudder.io"
@@ -34,21 +32,23 @@ tags:
   - name: Remote run
     description: Trigger agents runs
 paths:
-  "/system/status":
+  "/rudder/relay-api/1/system/status":
     $ref: paths/system/status.yml
-  "/system/info":
+  "/rudder/relay-api/1/system/info":
     $ref: paths/system/info.yml
-  "/system/reload":
+  "/rudder/relay-api/1/system/reload":
     $ref: paths/system/reload.yml
-  "/shared-folder/{path}":
+  "/rudder/relay-api/1/shared-folder/{path}":
     $ref: paths/shared-folder.yml
-  "/shared-files/{targetNodeId}/{sourceNodeId}/{fileId}":
+  "/rudder/relay-api/1/shared-files/{targetNodeId}/{sourceNodeId}/{fileId}":
     $ref: paths/shared-files.yml
-  "/policies/{nodeId}/rules/dsc/rudder.zip":
+  "/rudder/relay-api/1/policies/{nodeId}/rules/dsc/rudder.zip":
     $ref: paths/policies.yml
-  "/remote-run/nodes/{nodeId}":
+  "/rudder/relay-api/1/remote-run/nodes/{nodeId}":
     $ref: paths/remote-run/node.yml
-  "/remote-run/nodes":
+  "/rudder/relay-api/1/remote-run/nodes":
     $ref: paths/remote-run/nodes.yml
-  "/remote-run/all":
+  "/rudder/relay-api/1/remote-run/all":
     $ref: paths/remote-run/all.yml
+  "/metrics":
+    $ref: paths/metrics.yml

relay/sources/api-doc/paths/metrics.yml

@@ -0,0 +1,72 @@
+# SPDX-License-Identifier: CC-BY-SA-2.0
+# SPDX-FileCopyrightText: 2013-2020 Normation SAS
+get:
+  summary: Get metrics
+  description: Get metrics in prometheus format. You need to configure prometheus to scrape `http://localhost:3030/metrics` on the relay.
+  operationId: getMetrics
+  responses:
+    "200":
+      description: Prometheus metrics
+      content:
+        text/plain:
+          schema:
+            type: string
+            format: Prometheus metrics
+            example: |
+              # HELP rudder_relayd_inventories_total Agent inventories
+              # TYPE rudder_relayd_inventories_total counter
+              rudder_relayd_inventories_total{status="forward_error"} 0
+              rudder_relayd_inventories_total{status="forward_ok"} 0
+              # HELP rudder_relayd_managed_nodes_total Managed Nodes
+              # TYPE rudder_relayd_managed_nodes_total gauge
+              rudder_relayd_managed_nodes_total 0
+              # HELP rudder_relayd_reports_processing_duration_seconds Reports processing
+              # TYPE rudder_relayd_reports_processing_duration_seconds histogram
+              rudder_relayd_reports_processing_duration_seconds_bucket{le="0.005"} 1
+              rudder_relayd_reports_processing_duration_seconds_bucket{le="0.01"} 1
+              rudder_relayd_reports_processing_duration_seconds_bucket{le="0.025"} 2
+              rudder_relayd_reports_processing_duration_seconds_bucket{le="0.05"} 3
+              rudder_relayd_reports_processing_duration_seconds_bucket{le="0.1"} 3
+              rudder_relayd_reports_processing_duration_seconds_bucket{le="0.25"} 3
+              rudder_relayd_reports_processing_duration_seconds_bucket{le="0.5"} 3
+              rudder_relayd_reports_processing_duration_seconds_bucket{le="1"} 3
+              rudder_relayd_reports_processing_duration_seconds_bucket{le="2.5"} 3
+              rudder_relayd_reports_processing_duration_seconds_bucket{le="5"} 3
+              rudder_relayd_reports_processing_duration_seconds_bucket{le="10"} 3
+              rudder_relayd_reports_processing_duration_seconds_bucket{le="+Inf"} 3
+              rudder_relayd_reports_processing_duration_seconds_sum 0.043532457999999996
+              rudder_relayd_reports_processing_duration_seconds_count 3
+              # HELP rudder_relayd_reports_total Agent run reports
+              # TYPE rudder_relayd_reports_total counter
+              rudder_relayd_reports_total{status="error"} 1
+              rudder_relayd_reports_total{status="forward_error"} 0
+              rudder_relayd_reports_total{status="forward_ok"} 0
+              rudder_relayd_reports_total{status="invalid"} 1
+              rudder_relayd_reports_total{status="ok"} 2
+              # HELP rudder_relayd_sub_nodes_total Nodes behind this policy server
+              # TYPE rudder_relayd_sub_nodes_total gauge
+              rudder_relayd_sub_nodes_total 0
+              # HELP process_cpu_seconds_total Total user and system CPU time spent in seconds.
+              # TYPE process_cpu_seconds_total counter
+              process_cpu_seconds_total 0.12
+              # HELP process_max_fds Maximum number of open file descriptors.
+              # TYPE process_max_fds gauge
+              process_max_fds 1024
+              # HELP process_open_fds Number of open file descriptors.
+              # TYPE process_open_fds gauge
+              process_open_fds 27
+              # HELP process_resident_memory_bytes Resident memory size in bytes.
+              # TYPE process_resident_memory_bytes gauge
+              process_resident_memory_bytes 29184000
+              # HELP process_start_time_seconds Start time of the process since unix epoch in seconds.
+              # TYPE process_start_time_seconds gauge
+              process_start_time_seconds 1609155927.83
+              # HELP process_virtual_memory_bytes Virtual memory size in bytes.
+              # TYPE process_virtual_memory_bytes gauge
+              process_virtual_memory_bytes 1491361792
+  tags:
+    - System
+  x-code-samples:
+    - lang: curl
+      source:
+        $ref: ../code_samples/curl/metrics.sh